DE19604795C2 - Color separation using multi-point narrow band illumination of light sources with N colors - Google Patents

Description

The present invention relates to optical color images probes in general and in particular on a Method and device for separating color image data from a scanned object.

Color optical scanners produce color image data gnale that represent an object or document that is carried out by an illustration of the object or the document projected onto an optical photosensor array. The color image data signals can then be digitized and used saved for later use. For example, NEN the color image data signals through a personal computer used to draw an image of the scanned object a suitable display device, for example one CRT (CRT = cathode ray tube) testify.

A typical optical scanner has illumination and optics systems to illuminate the object and a small area of the illuminated object, which is usually is referred to as a "scan line" on the photosensor focus array. The entire object is then removed by moving the illuminated scan line over the entire ob is moved by either the object relative to the Lighting and optics arrangements is moved, or by the lighting and optics arrangements relative to the Ob be moved.

A typical lighting system for a color scanner device can have a suitable white light source for example a fluorescent tube or an incandescent lamp to illuminate the object. A typical optical system can  ne lens arrangement to illuminate the image scanned line on the surface of the optical photosen focus sorarrays, and may further include one or more Have mirrors to "fold" the path of the light beam, which enables the optics system to be conveniently combined in one relatively small housing is attached.

Although different types of photosensor devices ver can be turned to the light from the illuminated Ab Detecting the tactile line is a commonly used one Sensor the charge-coupled device or CCD (CCD- charge coupled device). A typical CCD can be a big one Have number of individual cells or "pixels" of which each in response to exposure to light an elec accumulates or builds up. Because the size of the battery mulated electrical charge in any given cell or each pixel on the intensity and duration of the exposure , a CCD can be used to make bright and dark spots on an image that focuses on the same is to grasp. In a typical scanner arrangement The charge buildup in each of the CCD cells or -Pixel measured and then at regular intervals are known as sampling intervals, which discharge for a typical scanner about five milliseconds or so can be the same.

Color optical scanners of the type described above are typically effective by collecting multiple color component images of the object being scanned. For example, data representing red, green, and blue color components of the scan line image may be generated, correlated, and stored by the scanner. The individual color components, for example red, green and blue, are usually referred to as primary colors, primary stimuli (primary stimuli) or simply "primaries" in English. As is well known, various combinations of three such primary colors can be used to create any color stimulus included on the CIE color chart in the spectrum of colors that lie in the triangle of primary colors. The amounts of the primary color that are required to match a specific color stimulus are called tristimulus values. Expressed mathematically:

C ∼ r (R) + g (G) + b (B)

In other words, that means a given color stimulus (Color stimulus) C (e.g. the image of the scan line) by r units of the primary stimulus R (red), g units of the Primary stimuli G (green) and b units of primary stimulus B (blue) can be tuned. All different physi calic stimuli that are just like the given color stimulus C the same three tristimulus values r, g and b have. Consequently, it is possible to create a color stimulus by a mixture of the three primary colors or stimuli agree, with the tristimulus values r, g and b the required determine the amount of each primary color. It is important tig to remember that the aforementioned Ver drive only achieved a psychophysical color adjustment (i.e. the color becomes the same for the human eye seem), as opposed to a physical or spec central adjustment.

Many different techniques have been developed to collect data collect, the multicolor component images (i.e. the Tristi mulus values) of the object being scanned. One technique is to image the illuminated ab tracing line towards a single linear photosensor array project. However, the multi-color component images (i.e. the tristimulus values) of the illuminated scan line collect a light source of different colors (one Primary color) used to scan the line at each of to illuminate three sweeps. For example the object first using only red Light are scanned, then only green light and finally only blue light. The exit  signal from the photosensor for each color thus represents represents the tristimulus value for this color. In a deviation This technique can take three sweeps under Ver using a white light source, where the light from the illuminated scan line during each of the three passes, however, through a different color film ter is filtered before it reaches the optical photosensor ray is focused. In any case, the tristimulus Values for the primary colors (i.e. the colors red, green and blue) who determines from the output signal of the photosensor the.

Another technique described in U.S. Patent 4,709,144 and U.S. Patent 4,926,041, which are hereby incorporated by reference, is to divide the illuminated (ie, multi-color) scan line into multi-color components, each of which is then radiated is focused on multiple linear photosensor arrays. For example, the illuminated scan line can be divided into red, green, and blue color component parts that are then projected simultaneously onto three ( 3 ) separate linear photosensor arrays. The output from each photosensor represents the tristimulus value for the corresponding primary color. This technique enables the tristimulus values for any single scan line to be generated simultaneously, allowing easier correlation of the image data for each separate primary color.

Regardless of the particular technique used to collect the tristimulus values, the color accuracy of the reproduced image will only be as good as the spectral band adjustment between the spectral sensitivity of the photo sensor used to record the image and the spectral sensitivity of human eyesight. As is well known, the human eye has three different types of color receptors (koni) that are sensitive to different spectral bands or regions that roughly correspond to red, green and blue light. The receptors are relatively "broadband" devices that are sensitive to a wide range of wavelengths in each ribbon region, as shown in FIG. 1. For example, blue receptors are typically sensitive to light that has wavelengths in a range from about 400 nm to 500 nm; Green receptors on light that has wavelengths in a range from about 480 nm to 600 nm; and red receptors on light having wavelengths in a range from about 500 nm to 650 nm. Although the specific sensitivities of the color receptors vary from person to person, the average response for each receptor has been quantified and is known as the "CIE standard observer"("CIE standard observer").

The ability to accurately reproduce colors in a tristimulus system, at least on a psychophysical basis, requires a fairly accurate spectral band match between the primary colors and the receptors. Therefore, an accurate color reproduction can be obtained by ensuring that the primary colors have spectral bands or regions that are as close as possible to the spectral response regions of the three receptors (ie red, green and blue) in the human eye to match. Since the receptors of the human eye are sensitive to a relatively wide spectral range of light colors (see FIG. 1), the filters or illumination sources used in color scanning devices must have similarly wide spectral ranges if an accurate color reproduction is to be obtained . Unfortunately, it is difficult to produce light sources and / or filters with broad-band spectral ranges that come very close to those of human receptors, let alone create an identical match. In addition, the demand for such broadband devices limits the number of alternative devices available to the developer, thereby limiting the ability to produce smaller, lighter, or cheaper color scanner designs. An even more significant disadvantage is that such broadband devices tend to limit the signal-to-noise ratio of the scanner, which can result in poorer image quality.

Accordingly, there remains a need for a color scanner device that does not depend on broadband light sources or filters pending trying the spectral bands of the recepto of the human eye. Ideally should such a scanning device be able to relatively compact and inexpensive narrowband lighting to use directions, for example light emitting Diodes, but without the color fidelity or the Si Reduce signal / noise ratio of the color image data. On additional benefits could be realized if a sol che scanner further correct the color image data could be such that the object appears to be under ner different light source has been recorded. For example, it is often desirable to image the abgeta constant object on a computer display device, übli a CRT to reproduce. If the light source, for which the color scanner is calibrated, from the equivalent light source of the computer display device differs, then a color shift is perceived become. In the same way, if the display device is a color printer, the light source for which the printer is calibrated, identical to the scanner light source calibration to verify a color shift avoid.  

DE 41 19 489 discloses a three-range process a flatbed scanner, which has a color separation with three Fluorescent lamps illuminating one after the other in the Red, green and blue areas reached. This flatbed scan ner does not include a correction filter and has only one CCD line le, where the individual picture lines correspond to each other in the ro emitting in the green and blue spectral range Light source to be exposed. Basically it turns into color analysis according to the three-area method requires that the effective spectral channel sensitivities of the flatbed scan with the three spectral sensitivities of the cones of the human eye in a suitable manner. For technical flatbed scanners, however, it is cheaper, not to strive for the basic spectral value curves, but CIE spec tralwertkurven that over a fixed coefficient matrix with are linked to the basic spectral value curves. Alternatively exi also a fixed conversion matrix to spectral value curves according to the EBU receiver primary colors in the Convert basic spectral value curves.

U.S. Patent No. 4,930,008 discloses a color image scanner with a plurality of light sources that light a first emitting diode for emitting a red light, one second light-emitting diode for emitting a yellow one Light and a fluorescent light source for emitting one blue light. The color image scanner also includes a control circuit for switching the plurality of lights swell in a sequential manner and an image sensor for receiving light reflecting from the scanned image is tiert. The color characteristics of the LED and the floures zenzlichtquelle are chosen so that a wavelength over lapping with the emission light of the yellow-emitting LED is available.

It is the object of the present invention, a method to collect color image data from an object, with the compact and affordable narrowband lighting system directions are used with color fidelity and the signal-to-noise ratio of the color image data is not correct is induced.

This object is achieved by a method according to claim 1 solved.  

It is another object of the present invention Color scanners to create the color image data without a reduced color fidelity or a reduced Si Signal / noise ratio with relatively compact and inexpensive Generate narrow band lighting devices.

This task is accomplished by color scanners according to the Claims 5 and 8 solved.

A color scanner using a color bar multi-point narrow-band lighting from Schmal band light sources with N colors, for example LEDs a multi-color light source for successive lighting of an object with light of N different colors, for example between a color red and a color blue. On Photosensor that responds to light from the object right is reflected, the reflected light converts into a sentence of color image data signals that each of the N different colors. Color image data processing The direction connected to the photosensor changes the Sets of color image data signals that each of the N below correspond to different colors, in equivalent tristimu lus values.

A method of color scanning can include the following steps indicate: consecutively illuminating the object Light of different colors that go beyond the visible Spectrum are distributed, d. H. about between a color red and a color blue; Collect corresponding N color image data of light shining through the object during the time that the Object successively by each light of the N colors is lit, is reflected; and converting the N colors image data in equivalent tristimulus values.

Advantageously, the primary colors that make up the Tristi correspond to mulus values into which the color image data are converted delt be selected to any desired  To be color or a spectrum of colors, although it is usually it will be desirable to choose primary colors that match the color matching functions of the CIE standard observers speak. Accordingly, the present invention is in La to generate color image data that exactly matches the color of the represent palpated object, but not on Light sources or filters with spectral energy distributions too must resort to the receptors of the human eye correspond. In a preferred embodiment, NEN narrow-band light sources, such as LEDs, are used, which are not only cost savings but also increase design flexibility, in that LEDs facilitate the development of compact light, inexpensive light sources enable order. Another advantage of the LEDs is that their relatively narrowband Light output the overall signal / noise ratio of the color image data increased, which increases the overall image quality becomes.

Preferred embodiments of the present invention are referred to below with reference to the attached drawing nations explained in more detail. Show it:

Fig. 1 is a graph showing the normalized spectral sensitivities of the three types of receptors in the human retina;

Fig. 2 is a perspective view of a game Ausführungsbei an optical scanner using a color separation with a multi-point narrowband illumination of LEDs with four colors;

Fig. 3 is a perspective view of the optical pickup device of Fig. 2 with the top plate removed;

Fig. 4 is a cross-sectional elevation of the carriage arrangement of the optical scanning device of Figs. 2 and 3;

Fig. 5 is a plan view of a photosensor array used in the optical pickup;

Fig. 6 is a top view of a document being scanned by an optical scanner showing the movement of an illuminated scan line across the document;

Fig. 7 is a block diagram of a control and Datenverarbei tung system for converting color image data into sets of tristimulus values;

Fig. 8 is a perspective view of an exemplary embodiment of a game four-color LED arrays for sequentially illuminating the object being scanned;

Fig. 9 is a perspective view of another embodiment of an exporting approximately four-color LED arrays to successive illumination of the object that is abge samples; and

Fig. 10 is a flowchart showing the method of generating color image data using a method of main component analysis.

Color separation using multi-point narrowband illumination of N-band narrowband light sources, e.g. LEDs, is shown in Figs. 2 through 5 as it could be used on an optical color scanner 10 to produce machine-readable color image data that a Represent a color image of an object 12 , for example a sheet of paper with colored graphics provided thereon, as shown in FIG. 6. Essentially, the optical color scanner 10 may include a housing 22 having an upper plate 14 on which a transparent platen 16 is mounted to support the object 12 (not shown in Fig. 1) to be scanned. The scanning device 10 also has a carriage assembly 18 which is adapted to be moved back and forth below the transparent support plate 16 along the scanning axis AA in order to achieve the function of scanning the object 12 .

As shown in FIGS. 3 to 5, the housing 22 of the optical scanning device 10 may have a suitable displacement arrangement 20 , on which the carriage arrangement 18 is attached. The displacement arrangement 20 moves the Wagenan arrangement 18 below the transparent platen 16 backwards and forwards. The carriage assembly 18 may include a housing 24 that is adapted to accommodate a multi-color light source assembly 26 , an optional slot opening assembly 28 , as well as the various components of the optics system 30 that include first, second, and third mirrors 32 , 34 and 36 and a lens arrangement 38 for focusing light rays that are never reflected from the scanning line 42 ( FIG. 6) onto the photosensitive surface 64 of a suitable photosensor, for example a CCD array 62 ( FIG. 5) which in the detector arrangement 44 is included. The scanning device 10 may further comprise a suitable electronic control and data processing system 70 ( FIG. 7) for controlling the various components and functions of the scanning device 10 and for processing the color image data from the detector arrangement 44 . Alternatively, a suitable data processing device, such as a personal computer (not shown) connected to the scanner 10 , may be used to process the color image data.

Briefly, the electronic control and data processing system 70 shown in FIG. 7 has a central processing unit or CPU 74 (CPU) that controls the photosensor or CCD array 62 , a programmable only Read memory or PROM 76 (PROM = programmable read-only memory), an analog / digital converter (A / D) 78 and a storage device, for example a random access memory (RAM) 72 (RAM = random access memory) . The PROM 76 can be programmed to use the principal component analysis method to convert the sets of color image data corresponding to the N colors used to illuminate the object to the corresponding tristimulus values, as explained in more detail below becomes. The tristimulus color image data can then be stored in RAM 72 , which can be accessed by a suitable optical data system (not shown).

The multi-color light source assembly 26 is capable of selectively generating light of N colors in a range across the visible spectrum from about one color blue to about one color red. In a preferred exemplary embodiment, the light source arrangement 26 can have an LED array 46 (LED = Light Emitting Diode) for generating four different illumination colors which roughly correspond to the colors blue, blue green, green and red. The LED array has a plurality of separate light emitting elements 48 ( FIGS. 3, 8) of the four different colors. The separate light-emitting elements 48 are connected in such a way that all elements 48 can be illuminated simultaneously for a given color, whereby the object that is scanned is illuminated with the individual color.

The object 12 is scanned by operating the displacement assembly 20 to move the carriage assembly 18 along the scan axis AA ( FIG. 3), causing the illuminated scan line 42 to be in a direction generally indicated by an arrow 50 is, as best seen in FIG. 6, over which object 12 is moved. Instead of illuminating the object 12 with white light, as is done with some color scanners, however, the object 12 scanned by the scanner 10 is sequentially arranged with light from each of the four ( 4 ) different colors from the multicolor light source voltage 26 illuminated. That is, the scan line 42 is moved four times over the object 12 , the object 12 being illuminated with a different color each time. During each sweep, color data relating to the particular color of the illuminating light is collected by the photosensor or CCD array 62 ( FIG. 5) included in the detector array 44 ( FIG. 4). After the four runs are completed, the RAM 72 in the control and data processing system 70 contains sets of color image data corresponding to each of the four colors of lighting. Control and data processing system 70 then uses the method of principal components to convert the sets of color image data into tristimulus values that correspond to a desired set of primary stimuli.

As explained in more detail below, the main component method is based on the assumption that if the spectral reflectivities of the object are known, the colors of the object will be known under all possible lighting conditions. The first step in the main component process is to find a small number of main components, the linear combination of which approximates the reflection spectra of the object. Once the main components of the reflection spectra of the object are determined, the tristimulus values of the object under the recording light source (e.g., multi-color light source assembly 26 ) can be corrected to a desired set of primary stimuli using the appropriate transfer function.

In a preferred embodiment, the one desired te set of primary stimuli to be selected to those Pri to be marvels that have a color match with the CIE Generate standard observers. The resulting tri stimulus values then become the colors of the sampled Ob  project exactly, even if the four colors that are used to illuminate the object, no spectral showed energy distributions with spectral energy distributions of the receptors of the human eye voices.

There is a significant advantage of the present invention then in that it is capable of displaying color image data generate exactly the color of the scanned object set, but without back on light sources or filters to have to grasp the spectral energy distributions with those of the receptors of the human eye to match. In fact, with a preferred one Embodiment relatively narrow-band light sources, at for example LEDs. The ability to be more common Replacing lighting sources with LEDs is not only a problem Represents cost savings, but also increases the design flexibility in that LEDs make the design easier compact, lightweight and inexpensive light source enable arrangements. Another benefit of having is linked to the ability to use LEDs rin that their relatively narrow-band light output the Ge Velvet signal / noise ratio of the color image data increases where is increased by the overall image quality.

Color separation under Ver multi-point narrowband illumination of LEDs with N Colors realized according to the present invention. Even though a preferred embodiment, for example four un Different colors of LEDs (blue, blue-green, green and red) can use any number of colors be applied, which is generally an increased color rendering loyalty with increasing numbers of colors. Whether probably the main component method is the color image data of the N Colors in corresponding tristimulus values (i.e. color image da of three colors) for a set of primary colors that Have spectral energy distributions that those of the recipe gates of the human eye, can convert,  the procedure is not based on any particular set of Primary colors are limited and can be used to correct the ent speaking tristimulus values for each desired set of To generate primary colors. The main component process can can also be used to correct the color image data such that the object appears to be under one different light source was recorded.

The details of an embodiment of an optical pickup 10 that uses color separation using multi-point narrowband illumination of LEDs with N colors are best understood by referencing FIGS. 2-4 with an occasional reference to FIG. 5 at the same time . 6 and 7 can be seen. The optical Tastvor direction can essentially have a housing 22 with an obe ren plate 14 which is adapted to receive a transparent platen 16 . The object 12 ( Fig. 6) can be placed face down on the transparent platen 16 to achieve the scanning function. The carriage assembly 18 is slidably secured in the scanner housing 22 and is moved back and forth below the transparent platen 16 by the displacement assembly 20 .

As shown in FIGS. 3 and 4, the carriage assembly 18 may include a trolley body having 24, which is adapted to the multi-color light source assembly 26 and the Schlitzöffnungsan order to take 28, although a slot opening assembly 28 is not always necessary. The carriage assembly 18 further includes the various components of the optical system 30 . The carriage assembly 18 is attached to the scanner housing 22 and the displacement assembly 20 so that the same can be moved relative to the transparent platen 16 and the object 12 carried thereon in a manner well known in the art may to generate a swept scan image of object 12 at image region 60 over photosensitive surface 64 of photosensor or CCD array 62 ( FIG. 5). Generally speaking, displacement assembly 20 may include a drive motor 52 and a drive belt 54 sen to move back and forth along the scanning axis AA below the platen. However, since many different types of shift arrangements 20 well known to achieve the aforementioned features and professionals are available, the Verschiebungsanord is planning 20 not described in detail.

Optical system 30 , which is mounted in carriage assembly 18, may include a plurality of mirrors 32 , 34 and 36 mounted in housing 24 to define a folded light path 56 that extends from the illuminated scan line 42 ( FIG. 6) of the object 12 extends through an opening 56 and a lens arrangement 38 onto the photosensor or the detector arrangement 44 . Again, since carriage assemblies of the type described above are well known to those skilled in the art, the particular carriage assembly 18 used in a preferred embodiment of the scanner 10 will not be described in greater detail. However, the structure of the optical pickup 10 may, for example, be similar or identical to the optical pickup device described in U.S. Patent 4,926,041, which is hereby incorporated by reference.

It should be noted that although the scanner 10 shown and described herein, scanning is accomplished by moving the light source and optical assemblies 26 and 30 (ie, carriage assembly 18 ) relative to object 12 , the same Scanning could be accomplished just as easily by moving the object relative to stationary lighting and optics arrangements as would be apparent to those skilled in the art. Such an alternative embodiment would include a stationary light source and stationary optics assemblies (not shown) and a movable platen (not shown) to move the object being scanned relative to the fixed light source and the most rigid optics assemblies.

The electronic control and data processing system 70 for collecting the sets of color image data corresponding to the N colors of the lighting and transforming them into desired sets of tristimulus values is shown in Fig. 7 and has an analog-to-digital converter 78 (A / D), which is connected to the data bus 80 of a photosensor or CCD array 62 . The A / D converter 78 operates in a conventional manner to convert the analog signals received from the data bus 80 of the CCD array 62 to digital signals that are stored in a storage device, such as a random access memory ( RAM) 72 can be saved. As described above, the data signals received by the photosensor or CCD array 62 correspond to sets of color image data signals corresponding to each of the N different colors of the illuminating light from the multicolor light source assembly 26 . The control system 70 includes a programmable read only memory (PROM) 76 which is connected to the CPU 74 and the RAM 72 and which can be programmed to use the method of the main components to generate the sets of color image data, convert stored in RAM 72 to tristimulus color data as described below. CPU 74 , which is connected to photosensor or CCD array 62 , PROM 76 and RAM 72 , controls the operation thereof.

The details of the multi-color light source assembly 26 are best seen in FIGS. 4 and 8. Essentially, the light source assembly 26 may include an LED array 46 that is attached to the slot opening assembly 28 on either side of the opening 56 that defines the scan line, although a pair of LED arrays located on either side of the opening 56 are attached, could also be used. The LED array 46 has a plurality of individual diodes or light-emitting elements 48 which are attached to a suitable base or substrate arrangement 68 . Each light emitting element 48 is designed to produce light of a different color. For example, in the exemplary embodiment shown in FIG. 7, the LED array 46 generates light of four different colors, which roughly correspond to the colors blue, blue-green, green and red. Therefore, each LED array 46 has four different types of light emitting elements 48 , each of which corresponds to a specific color. That is, the light emitting elements 48 may include a plurality of blue elements B, blue-green elements BG, green elements G, and red elements R. In one embodiment, the light emitting elements of the four different colors can be arranged sequentially, starting at one end 66 of the array 48 with a blue element B, then a blue-green element BG, a green element G, a red element. Element R and then repeats the sequence along the entire length of array 46 .

Another embodiment of the array 46 may use different numbers of each of the four colors of the light emitting elements 48 , depending on the spectral output characteristics of the individual light emitting elements, the spectral response characteristics of the detector assembly 44 , or both. For example, if the spectral power output of the green elements G 48 is significantly less than the spectral power outputs of the other color elements, it may be necessary to include more green elements G 48 in the array 46 compared to the other color elements. Such an increase in the number of green elements G with respect to the other color elements, for example blue B, blue-green BG and red R, will help to make the output signals from the detector arrangement 44 the same for each of the four colors of light sources. If the spectral response of the detector array 44 is lower for a certain light color than for others, it may likewise be necessary to increase the number of light-emitting elements 48 for the corresponding color in order to compensate for the deficit. In any event, it desires that the output signals from detector array 44 should be substantially the same for each of the four colors of light sources, e.g., blue B, blue-green BG, green G, and red R, if a white one Object 12 is scanned. Therefore, the numbers of all four different color elements 48 , e.g., blue B, blue-green BG, green G, and red R, should be varied, if necessary, to obtain output signals of nearly equal strength from the detector array 44 , if a white one Object is scanned.

Regardless of the numbers of all four different color elements 48 used in the array 46 , all light emitting elements 48 associated with a single color are connected to the control and data processing system 70 of the scanner 10 such that they are all the same can be switched on or off simultaneously and independently of the others. For example, if all of the blue B light emitting elements 48 are turned on, the LED array 46 will emit blue light. When the blue B light emitting elements 48 are turned off and the blue-green BG elements 48 are turned on, the LED array will emit blue-green light.

The colors of the light emitting elements 48 Kgs nen in any suitable manner on the sub strate be arranged 48th For example, a further exemplary embodiment of an LED array 146 is shown in FIG. 9, in which the individual colors blue B, blue-green BG, green G and red R of the light-emitting elements 148 are arranged on a substrate 168 such that all the elements 148 of the same color form lines that extend from a proximal end 166 to a distal end 170 of the array 146 . This means that all blue elements B 148 contain a first line; all blue-green elements BG a second line; all green elements G a third line; and all red elements R a fourth line. As with the first embodiment of the LED array 46 shown in FIG. 8, all of the light emitting elements 148 of the same color are connected so that they can all be turned on or off together, thereby enabling that Array 146 selectively emits light of four different colors.

The method 82 for generating color image data corresponding to the four different colors of light generated by the multi-color light source arrangement 26 is shown in FIG. 10. In a first step 84 , the CPU 74 switches on a first color element 48 on the LED array 46 , for example blue. In a step 85 the object is then scanned by actuating the displacement arrangement 20 in order to move the carriage arrangement 18 below the support plate 16 such that the blue illuminated scanning line 42 extends over the entire object 12 along the scanning direction 50 (see Fig. 6) moves. The output data signals from the CCD 62 , which correspond to the blue light illumination, are then collected and stored in the RAM 72 of the electronic control and data processing system 70 ( FIG. 7) in a step 86 . Alternatively, the color image data signals can be stored in a memory system that is included in an external processing system, such as a personal computer (not shown).

After the scanning is complete, the CPU performs a step 87 to determine whether the object 12 has been scanned with all colors. If not, the CPU executes step 88 to turn off the last color (blue) and turn on the next color (e.g. blue-green). Thereafter, step 85 is repeated, except that the object 12 is scanned under the blue-green light source. The image data from the photosensor array 62 are again collected and stored in step 86 , this time as blue-green image data. The aforementioned scanning process is repeated for each of the N colors (four colors in this case) until the object 12 has been scanned with each color. After all scans, RAM 72 will contain N sets of color image data corresponding to all of the N color light sources. More specifically, in the embodiment just described, RAM 72 will contain four separate sets of color image data - blue, blue-green, green, and red color image data.

Once the object 12 has been scanned among all colors, as determined in step 87 , the CPU 74 next performs a step 89 and uses the principal component analysis method to determine the color image data stored in RAM 72 (ie, the color image data, which correspond to the N colors of the illumination) to be converted into tristimulus image data. Advantageously, the set of primary colors corresponding to the tristimulus image data can be selected to be any desired color or color spectrum, although it will usually be desirable to select a set of primary colors that have spectral energy distributions that match them correspond to the receptors of the human eye as described above. That is, the tristimulus values into which the image data is transformed will be the color matching functions for the standard CIE observer.

Although the process of main component analysis to professionals is well known and in each of a number of sources too can be found, for example, with M. J. Vrhel and H. J. Trussell, "Color Correction Using Principle Components", Color Re search and Applications Journal, Vol. 17, No. 5, October 1992, pp. 328 to 338, this document hereby by Be access, this procedure becomes the purpose creating a basis for understanding the present Invention briefly described herein.

Essentially, the process is the main component ana lyse a mathematical procedure that is suitable for analysis of color systems. When applied to color systems the main component analysis process is based on assuming that if the spectral reflectivity in the colors of the surfaces in the image are known Image known under all possible lighting conditions  will be. A standard assumption when evaluating the spectral reflectivity is that the reflectivity spec tren adequately by a linear combination of a small An number of main components can be approximated. Nearly all printed color images, which are the objects, the usual can be scanned using the Four-color printing process generated. With the four-color printing all colors in the picture will be driven by the application different proportions of the three colors of the printing ink te, yellow, magenta and cyan, reproduced, black in Shadow areas is added. Since the spectral reflect Xions ability of the printing inks commonly used and pigments are well known, and because of the spectral energy distribution of the different colors of LEDs can be, and consequently known, the method of the main components used to measure the spectral Re to approximately determine the inflection capacity of the image and transform the N color image data into tristimulus values, which match the primary colors with any desired spec equal energy distribution.

Claims (13)

1. A method of collecting color image data from an object ( 12 ) comprising the following steps:
successively illuminating the object ( 12 ) with light of N colors, each light of the N colors having a different color between approximately a color red and a color blue;
Collecting corresponding N color image data of light reflected by the object ( 12 ) during the time that the object is successively illuminated by light of each of the N colors; and
Converting the N color image data into equivalent tristimulus values using principal component analysis, each equivalent tristimulus value having a primary stimulus with a desired spectral energy distribution. 2. The method according to claim 1, wherein the spectral energy Distribution of each primary stimulus essentially with a spectral energy distribution of a CIE standard Observers matches. 3. The method of claim 1 or 2, wherein N is four. 4. The method according to any one of claims 1 to 3, in which the four colors of light essentially about one Color blue, about a color blue-green, about a color be green and approximately one color red. 5. Color scanner ( 10 ) with the following features:
a multi-color light source ( 26 ) for sequentially illuminating an object ( 12 ) with light of N colors, each light of the N colors having a different color between about a color red and a color blue;
a photosensor device ( 44 ) responsive to light reflected from the object ( 12 ) to convert the reflected light into sets of color image data signals corresponding to each of the N colors;
color image data processing means operatively connected to the photosensor device ( 44 ) and responsive to the sets of color image data signals generated thereby to convert the sets of color image data signals corresponding to each of the N colors into equivalents using principal component analysis Convert tristimulus values, each equivalent tristimulus value having a primary stimulus with a desired spectral energy distribution. The color scanner according to claim 5, further comprising an imaging device for focusing light from an illuminated scan line ( 42 ) on the object ( 12 ) onto an image region and for providing an image of the illuminated scan line ( 42 ) on the object ( 12 ) the image region, the photosensor device ( 44 ) being positioned in the image region. The color scanner according to claim 6, further comprising a shifting device ( 20 ) for generating a relative shift between the object ( 12 ) and the image forming device, and for generating a swept scan image of the object ( 12 ) in the image region gene. 8. Color scanning device with the following features:
a multi-color light source ( 26 ) for sequentially illuminating an object ( 12 ) with light of N colors, each light of the N colors having a different color between about a color red and a color blue;
an image forming device for focusing light from an illuminated scan line ( 42 ) on the object ( 12 ) onto an image region and for providing an image of the illuminated scan line ( 42 ) on the object at the image region;
a photosensor device ( 44 ) positioned in the image region to generate color image data signals representative of the intensity of the light incident thereon;
displacement means ( 20 ) for generating a relative displacement between the object ( 12 ) and the image forming device and for generating a continuous scan image of the object ( 12 ) in the image region; and
a color image data processing device ( 70 ) operatively connected to the photosensor device ( 44 ), responsive to the sets of color image data signals generated by the same, to determine the sets of color image data signals corresponding to each of the N colors using principal component analysis convert to equivalent tristimulus values, each equivalent tristimulus value having a primary stimulus with a desired spectral energy distribution. 9. A color scanner according to any one of claims 5 to 8, wherein the multi-color light source ( 26 ) emits light of four different colors. 10. A color scanner according to claim 9, wherein the multicolor light source ( 26 ) comprises a plurality of light-emitting elements ( 48 ) for emitting the light of the four different colors, the light-emitting elements ( 48 ) being operatively connected to each other for each color, such that all light-emitting elements ( 48 ) can be operated independently for each of the four different colors. 11. A color scanning device according to claim 10, wherein the light-emitting elements ( 48 ) have light-emitting diodes (LEDs). 12. Color scanner according to one of claims 9 to 11, in which the four colors of light essentially about a color blue, about a color blue-green, about correspond to one color green and approximately one color red. 13. Color scanning device according to one of claims 5 to 12, wherein the photosensor device ( 44 ) has a linear CCD.